Challenges and insights from root segmentation in X-ray computed tomography data of Robinia pseudoacacia

1. Home
2. Publications
3. Online Library
4. Challenges and insights from root segmentation in X-ray computed tomography data of Robinia pseudoacacia

Robinia pseudoacacia is a nitrogen-fixing legume tree native to eastern North America, valued for its rapid growth, drought tolerance, durable timber, and suitability as a resilient urban tree. Through symbiosis relationship with Rhizobia bacteria, which reside in root nodules and convert atmospheric nitrogen into a plant-usable form, it thrives in nutrient-poor soils. However, little is known about how this symbiosis influences root system architecture (RSA). High-resolution X-ray computed tomography was used to investigate how RSA develops in response to symbiotic nitrogen fixation in nutrient-poor sand conditions. Robinia plants, genetically identical clones of Robinia, were initially grown in agar and later potted in poorly graded sand to enable the identification of individual grains. Rhizobia were introduced after initial growth to establish nitrogen-fixing symbiosis. By scanning the plants before and after inoculation, we can track the development of RSA.

A major challenge is image segmentation: root tissue and pore water exhibit similar X-ray attenuation, and nodules often fuse with fine root hairs. These factors hinder clean separation of the four present phases: Soil, root, water, and air. This study therefore focuses on developing segmentation workflows to achieve robust segmentation for further studies. 

Keywords: Soil vegetation; X-ray CT; Image analysis; Root morphology

3rd International Workshop on Soil-Vegetation-Atmosphere Interaction (RootS2025)

2a. Experimental laboratory characterisation of the multiscale bio-hydro-chemo-mechanical behaviour of rooted soils